AI doctors will be good at science but bad at business, and big talk with little action means even higher drugs prices: 10 healthcare predictions for 2026 from top investors

We nailed more payor CEOs losing their jobs (Andrew Witty and Lois Quam) and cell therapy coming back into fashion (four acquisitions, notably Capstan by Abbvie at $2.1B).  Otherwise, we were spectacularly poor. We were too skeptical about how long AI point solutions would remain healthcare investor darlings, propping up seemingly nonsensical businesses, and there was only one “down” health tech IPO, Hinge, that has done nicely since. We expected Medicare Advantage (MA) to come roaring back in 2025 but the Trump team has not put their thumb on the scales. We are surprised (and not sure we agree) that the Centers for Medicare and Medicaid Services (CMS) is forecasting MA to shrink in 2026 by 1 million members. We also expected to see risk adjustment and “Stars” eased and more generous premium increases. 

Big companies surprised us by what they did and didn’t do. Namely, HCSC’s willingness to incinerate a ton of money by acquiring Cigna’s MA book and Illumina, down 40% in 2025, has not been acquired (though nice Q3 earnings bounce). We were also surprised Apple did not launch glucose sensing with WatchOS 26, despite our recurring prediction that Big Tech will underdeliver in healthcare.  

After months of on-again, off-again tariffs, public health chaos, and sadly the cancellation of billions of research funding and healthcare coverage support for millions, we hope 2026 is calmer. Now on to our predictions for 2026 …

1. Humana crashes. Humana bid boldly for 2026, offering more generous MA plans in almost every county. We think this will result in extraordinary membership growth, north of 1M new lives. The problem with MA growth is that if you are not careful, you can lose a lot of money on new members. We think Humana is going to pile up huge losses and incite activist investors seeking change.
2. Large employers flee the big three Pharmacy Benefit Managers (PBMs). If we have learned one thing from the last several years of PBM oversight, it is that PBMs are great at inventing new fees and preserving their margins. We think large employers will finally say enough is enough with these shell games. Alternatives to the big three now have a track record of saving and service that makes switching worth the risk (FWIW, we predict that the performance of the alternative partners is going to go well).
3. CMS creates a billing code for AI. The Trump Administration is bullish on AI for everything in healthcare: reducing fraud, waste and abuse, data analytics, remote monitoring, reducing administrative complexity, preventive care, and expanding access to care. To encourage innovation, we think CMS will create billing codes for AI. This will spawn a slew of new products, some that gain exciting traction, and, unfortunately, also a lot of fraud. Also, we think it is far from certain if AI will make our healthcare cost problem better or worse. Without payment model innovation, we could see AI being used to optimize our existing fee-for-service model, which unfortunately could generate increased care and expense.  
4. Multiple Measles outbreaks. Sadly, we think the dismantling of the CDC coupled with many states making it easy to opt out of vaccines is going to go very badly.  Measles is the most infectious of childhood diseases and as community vaccine rates fall it will come roaring back.    
5. AI agents gone wild, but there are some benefits. Agentic AI is very rule-following and goal-oriented. We think this is going to lead to silliness. Soon, agents will be talking to agents and they will go in circles forever – not compromising and not saying no. While this will lead to the benefit of automating many phone and fax-based workflows, we expect there to be a massive number of escalations to humans without agreed upon shared rules among payors and providers. We also think that there will be a surge of complaints about agentic calls taking longer than old-fashioned calls done by humans.
6. Lower cost cell & gene therapies. One area where drug prices will fall dramatically over the next 5 years is cell and gene therapies. Breakthroughs in science and manufacturing will lead to $5,000-$50,000 COGS allowing lower prices to be passed on to payors. There will also be dramatically improved access to these treatments, not only as a result of decreased COGS but also workflows and toxicity profile improvements. We believe the first glimmers of this progress will be seen starting in 2026.
7. Big talk but no progress on commercial drug trend. Trump has talked a big game on drug prices — launching TrumpRx, cheap GLP-1s, tariff threats if U.S. prices don’t converge with the EU, and tougher Medicare negotiations. Despite the rhetoric, we expect the drug trend for employers and commercially insured patients to keep rising in 2026, not fall. Total drug spend will continue to climb as more patients initiate therapy, particularly with broader eligibility of GLP-1s as a result of the unit-price concessions provided by Novo and Lilly. The strategic challenge for plans and employers becomes not just paying for GLP-1s, but figuring out how to get patients off them by pairing therapy with meaningful lifestyle change to preserve the weight loss and health gains.  
8. Republicans keep control of the House and expand their majority. In a stunning upset, Trump avoids a midterm reset with redistricting, adding to the majority. We think this is made possible because the U.S. economy maintains its remarkable, AI-fueled buoyancy and Trump earns credit for the peaceful resolution of the Israel and Ukraine wars (we hope).  
9. Inflation runs hot. When we took economics, we learned that tariffs are taxes paid by consumers. We still believe that this is true. As a result, we think corporations cannot wait forever, despite legitimate fear of being singled out by a retribution-ready White House, to pass their costs on to consumers. We think we will see inflation running hot in 2026 and, as a result, interest rates will remain higher than expected as the Fed acts in defiance of Trump. While painful, we don’t think this will be enough to offset the forces which lead to our first prediction. 
10. AI doctors are good doctors and bad standalone businesses. We are stunned by how well AI can deliver clinical care. That said, we are not optimistic that AI-native clinical services will be durable standalone businesses. We do think these products and features will be awesome complements within a broader organization to deliver care, provide better access and make a lot more possible between visits, monitoring, and hopefully decreasing overall episode of care costs. As a result, we expect the traditional “iron triangle” of cost, quality, and access in healthcare to bend, as AI makes it possible to apply clinical intelligence at scale rather than only through human labor.  

We want to make two bonus predictions. Maybe this is overcompensation for our 2025 (under)performance, but we feel compelled to prognosticate that Judy Faulkner will decide that the risk of antitrust enforcement and being 82 makes it time to sell Epic to Microsoft, who knows a thing or two about antitrust, and thinks the gamble to gain the upper hand on Oracle is worth it.  

Finally, we are going to double down on our 2025 prediction that Apple does Glucose monitoring well. We have no insider knowledge, but we sure hope that this comes true since the public health benefit could be massive.   

We look forward to reporting back on these in a year and hope that our crystal ball proves clearer for 2026. We wish you all a joyful holiday season and healthful new year.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.